Getting Started

Install steps follow the official documentation. Commands assume a running Kubernetes cluster and Helm 3.

Prerequisites

• A Kubernetes cluster you can reach with kubectl.
• Helm 3 installed locally.

Install

KubeVela's core controller installs as a Helm chart from the project's chart repository (KubeVela docs):

helm repo add kubevela https://kubevela.github.io/charts
helm repo update
helm install --create-namespace -n vela-system kubevela kubevela/vela-core

A first working setup

After the controller is running, deliver an application by applying an Application custom resource with kubectl.

Write a minimal Application to app.yaml:

apiVersion: core.oam.dev/v1beta1
kind: Application
metadata:
  name: first-vela-app
spec:
  components:
    - name: express-server
      type: webservice
      properties:
        image: oamdev/hello-world
        ports:
          - port: 8000
            expose: true

Apply it:

kubectl apply -f app.yaml

The Application resource holds components, policies, and a workflow (src/apis/core.oam.dev/v1beta1/application_types.go:51-65). The webservice type and its properties are resolved by the matching ComponentDefinition at reconcile time.

Verify it works

Application resources are registered with print columns for phase and health, so a plain get shows status (src/apis/core.oam.dev/v1beta1/application_types.go:74-78):

kubectl get application first-vela-app

Wait for the PHASE column to read running. The controller-manager logs each reconcile as Start reconcile application ... End reconcile application (src/pkg/controller/core.oam.dev/v1beta1/application/application_controller.go:112-114), which is useful when a delivery is stuck.

Where to go next

For multi-cluster delivery, custom ComponentDefinitions and TraitDefinitions in CUE, addons, and production hardening, see the official documentation at kubevela.io/docs.